Title:

Author(s):

Submitted to:

COMPARISON WITH EUROPEAN OBSERVATIONS OF METEOR IMPACT

-

Gregory H. Canavan, DDP

For discussions outside the Laboratory

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsi- bility for the accuracy. completeness, or usefulness of any information, apparatus, product, or

en= herein to any specific commercial product, proctss, or service by trade name, trademark, manufacturer, or otherwise dots not necessarily constitute or imply its endorsement, ream- mendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

- - _ - - - process disclosed, or represents that its use would not infringe privately owned rights. Refer- - -

- -

. . ~ ~ -. .- - - ... --

Kp--- ..

Los Alamos N A T I O N A L L A B O R A T O R Y

Los Alamos National Laboratory, an affirmative actionlequal opportunity employer, is operated by the University of California for the U S. Department of Energy under contract W-7405-ENG-36. By acceptance of thts .Wide, the publisher recogntzes that the U S Government retains a nonexclusive. royalty-free Ifcense to publish or reproduce the published form of this contribution, or to allow others to do so, for U S Government purposes. The Los Alamos National Laboratory requests that the publisher identify this article as work performed under the auspices of the U.S. Department of Energy

Form No 836 R5 ST 2629 1 OB1

Portions of this document may be iliegible in electronic image products. Images are produced from the best available original document.

COMPARISON WITH EUROPEAN OBSERVATIONS OF METEOR IMPACT -.

Gregory H. Canavan

A model for the inference of object size and speed from observations is used to discuss European observations of impact. It compares the observed and predicted breakup altitudes for the objects larger than one meter and observes useful correlations. Trends in magnitude correlate well with measured velocities, altitudes, and trajectories and inferred size and strength parameters, but each parameter is subject to dispute, which can only be addressed when the sensitivity of predictions to uncertainties in these parameters is assessed.

Earlier notes19293 derived a model for the hydrodynamics, ablation, and radiation of meteor impacts at the level needed to infer meteor parameters from observations. The full model is presented in "Fragmentation and Ablation During Entry,"4 which is reviewed briefly below. The previous note5 compared it to other U.S. models with different but related features and objectives. The most closely related model is that due to Hills and Goda,6 which solves the same equations numerically, uses a consistent treatment of fragmentation, and produces similar results for common parameters, although it does not treat radiation in detail and assumes constant heat transfer coefficients. A model due to Chyba, Thomas, and Zahnle7 differs primarily in its estimate of fragment spread velocities and variable heat transfer coefficients. Both of those models are primarily concerned with interpreting the large Tunguska impactor, so they simplify many features and concentrate on varying parameters such as heat of vaporization and material strength to try to match observed surface damage patterns. The previous note discussed models with zero-dimensional hydrodynamics, simplified radiation transport, constant heat transport, parametric fracture and spread, and limited recourse to the limited data on objects tens to thousands of meters across.

This note extends that discussion to European data and models of meteor impact. U.S. analyses have used simplified models to try to achieve agreement with the very limited data on large objects while the European efforts have used more complex models that use observational data more directly to achieve agreement for a larger body of data on smaller objects. While the former is at a level of simplification better suited to the model for data inversion needed for this effort, the latter represent a more thorough test of model sensitivities and uncertainties for objects in the size range of interest. Thus, it is also a useful test of the inversion model. In cases of mutual applicability, the U.S. and European models appear to give broadly consistent results.

This note puts the inversion model into a form appropriate for interpretation of the European and Prairie Network observations of 14 objects larger than 1 m in diameter. The EN and PN observations use largely consistent classifications of these objects into stony,

1

carbonaceous, and cometary materials and indicate that most objects larger than 1 m are expected to be stony or cometary. When these objects are displayed according to their observed altitude of maximum brightness, which is related to fragmentation altitude, and inferred strength in bars, they fall into three groups: one stony object, the carbonaceous objects, and the soft cometary objects. Ideally, they should fall along the theoretical line of pressure versus altitude; in practice they exhibit several order of magnitude scatter due to variations in material strength. Correcting pressure for entry velocity provides a more precise measure of deviations from theory. There is a strong correlation, but there is a great deal of spread within each group.

Russian analysis program, but some qualitative observations are useful. The change in magnitude of the 14 large objects with size and speed can be explained readily by the changes in their size, speed, trajectory, and type. Trends in magnitude correlate reasonably well with measured velocities, altitudes, and trajectories and inferred size and strength parameters. However, each parameter is subject to dispute. The sizes are uncertain to factors of two; the masses to about an order of magnitude, and the strengths to a like amount. Thus, while the European and Prairie Network observations provide a useful data base for large objects, these uncertainties can only be addressed when the sensitivity of predictions to uncertainties in these parameters is assessed.

Review of model for inversion of observations. Earlier notes developed a model for objects that ablate and fragment during entry.8 That model also provides a common framework for the discussion of European efforts. The inversion model is formed by complementing the equation for the conservation of momentum

where M is the object mass, V is its instantaneous velocity, C is its drag coefficient, p is the local density, and A is the object area with an equation for the rate of change of mass due to ablation

where Q is the heat of vaporization and J is a heat transport coefficient, which is = 0.1 at altitudes 2 30 km. For constant J, taking the ratio of Eqs. (1) to (2) and integrating produces

where Mo and VO are the object's initial mass and velocity, v = VNo, and K = JVo2/2CQ is a parameter. The model is closed by assuming that fragmentation takes place when the ram pressure on the front of the object reaches its strength S , or

After the object breaks up, its fragments are assumed to move apart at a transverse velocity of

until they fall below the velocity of Eq. (4). Before fragmentation, the velocity can be approximated adequately by the form

The quantitative analysis of the EN-PN data is better discussed in conjunction with the

MdV/dt = - CpAV2, (1)

QdM/dt = - JpAV3, (2)

M = M&xp[-K(l - v2)], (3)

p = cpv2 , = s,

Vt = kVq(p/paD), (5)

(4)

2

v = voe-PWpCod, (6) where p = M/A = 0.7paD is the areal density (ballistic coefficient) of an object of mass M, area A, diameter D, and density ra and 8 is the angle of incidence with respect to vertical. After fragmentation, V can be described by a function that is more complicated, but which remains analytically soluble, differentiable, and invertible. Once V and M are known as functions of altitude, it is possible to invert the solutions to infer the initial Vo, Do, and angle of incidence from measured values of peak power, pulse duration, and altitude of maximum radiation.

a function of object parameters. The pressure on the object's front surface is given by Eq. (4), where to adequate accuracy the velocity can be approximated by Eq. (6). Thus, the pressure is

as shown in Fig. 1 as a function of altitude for vertical incidence at Vo = 20 km/s for D = 0.25 to 4 m. The names are observed objects discussed in the next section. The pressures for all objects are within a factor of two above 20 km, where they scale as p = CpVo2 - p. Because the pressure increases exponentially along with p, for large objects, the altitude of fragmentation is insensitive to the precise value of material strength, which is useful, as that parameter is poorly known. For larger objects, the peak pressure is larger at lower altitudes because they do not decelerate until p = - D. Smaller objects decelerate and produce lower pressures.

At 20 km, the pressure reaches about 200 bars (atmospheres) for 1-4 m objects and 100 bars for a 0.25 m object. Pressure scales as Vo2, so at 10 M s , the 4 m object would only reach 50 bars. In this approximation the angle of incidence enters only as Pcod - Dcos8, so non- vertical incidence at cos0 corresponds to vertical incidence of an object of diameter D/cose. At 25 km, 75 degree incidence (cos0 = 0.25) of a 1 m object would produce a pressure about 75/10 = 75% that from vertical incidence. For reference, the nominal yield stresses for iron, stone, carbon, and comet materials are 1,OOO, 100,10, and 1 bar.9 Thus, at 20 M s , iron objects larger than 2m should reach the ground intact. Stone objects of meter sizes would fragment at about 25 km. Carbonaceous objects would fragment at about 40 km. IIIA comets would fragment at about 60 km, and IIIB comets might fragment at about 70 km, as shown in Fig. 1.

European and Prairie Network (EN and PN) impact observations. This film- and video-based observation network has been in operation for a number of decades; it is the successor of the visual observation programs of the previous century. 10 Its advantage is its ability to triangulate positions and trajectories of objects and use established photometric correlations to estimate sizes; its limitation is its restricted geographic coverage. The networks have observed hundreds of objects, most much smaller than a meter. The observations of greatest interest here are those in Table I, which show the impacts of meteoroids larger than 1 m.l There

Entry pressures. A useful by-product of this model is a prediction of breakup altitude as

p = CpVo2e-2PWpCod, (7)

3

are a total of 14, ranging from the 10 m daylight Earth-grazing object of 10 August 1972 to the 1 m object of 14 August 1963. The interval spanned is from 1959 to 1994, about 35 years, so that the average impact rate is about 0.4/yr, although the rate is smaller for larger objects, as discussed further below.

For example, the 10 m daylight Earth-grazing object of 10 August 1972 row entries are its approximate 10 m diameter, -19 magnitude, 1,OOO tonne initial mass, maximum brightness altitude hM = 58 km, beginning luminous altitude h g = 76 km, ending luminous altitude hE = 102 km, angle of incidence cos0 = 0, initial velocity voo = 15 M s , and ending luminous velocity VE = 14.2 M s . The elements of orbit are of less concern, but a useful additional datum is the table inferred largely from observations of smaller objects12

Table 11. Survey of large meteoroid populations (source: Czplecha Table I) mx gssumed composition density ablation Strength 13 I stony 3.7g/cc .017 s 2 h 2 1OOO bar 11 carbonaceous 2 .041 100 IIIA cometary .75 .1 10 mB soft cometary -27 .2 1 l ?

This ablation parameter is related to the K = JVo2/2CQ of Eq. (3) above. For J = 2C and fixed VO, K - 1/Q, which for the Q = 8,5, and 2.5 MJkg used in the earlier notes for stony, carbonaceous, and cometary objects, respectively, give K's in the ratio of 1,2, and 4, in qualitative agreement with the progression seen in Table I1 for type I, 11, and IIIA objects, respectively, although the numerical values are normalized differently. The category IIIB, Le., very low density, easily ablated objects, were not treated in the inversion model. Note that Table I contains five IIIB objects, over 1/3rd of the total.

As an example, the 10 m daylight Earth-grazing object of 10 August 1972 is identified as a type I1 carbonaceous object. At 15 km/s, at 58 km it should produce a pressure of about 1 bar, which should not fragment a large carbonaceous object, although it may have in that its ending luminous altitude was higher than its beginning luminous altitude.

which gives the EN-PN estimates of the fraction of objects as a function of size. The fractions of IIIA and I objects fall from about 35 and 15% at 0.1 m to 15 and 10% at 1 m and 5 and 1% at 10 m, which means stony and cometary objects are a small minority at large sizes. Carbonaceous 11 objects are 30-45% throughout. IIIB increases from 20 to 60%. Based on this data, most objects larger than 1 m are expected to be stony or cometary, in agreement with Table 1.

Another important datum, also inferred from observations of smaller objects, is Fig. 2,

4

The objects from Table I are displayed on Fig. 1 according to their observed altitude of maximum brightness, which previous work has shown to be related to their fragmentation altitude, and their inferred strength in bars according to the I, 11, and IIIB classifications of Table 11. The objects fall into three groups: the stony Pribram by itself; the carbonaceous PN40503 through the Earth-grazing object; and the soft cometary Wurzburg through PN40401. Ideally, the objects should fall along the theoretical line from Eq. (7); in practice they exhibit several order of magnitude scatter. Pribram is far to the right and above the line, which indicates either an order of magnitude less strength than inferred or two orders of magnitude less penetration of density. While PN39470, PN39434, Freising, and Otterskir are close to the line, PN40503 and Beneslov are well below it, indicating greater smngth; and the Grand Prairie and Earth-grazing events are well above it, indicating less. In the cometary objects, PN40401 is the furthest from the curve, indicating less penetration ability than expected.

Impact pressure. The pressures in Fig. 1 are not corrected for entry velocity. Doing so provides a more precise measure of deviations from theory. For the entry velocities and brightness altitudes of Table I, Eq. (6) indicates that the velocity falls little by the altitudes of maximum brightness. Thus, the pressure depends little on size, and objects fragment at high altitudes where the pressure scales as p a CpV02. If an objec& tensile strength is given by S, it fragments at altitude where p = S/CV02, or

Figure 3 shows the observed fragmentation altitude ZO vs the predicted fragmentation altitude Zf for each of the objects in Table I, which can be identified by their altitudes of maximum brightness. There are now two groups of objects. The first, containing type I and II objects, is predicted to fragment at 35-45 km and is observed to fragment at 20-60 km. The second, containing IIIB objects, is predicted to fragment at 70-80 km and is observed to fragment at 60- 90 km. In each case, the variance is wider in the observed fragmentation altitudes than in the predicted ones due to flaws in the larger objects, which reduce their effective strength, as discussed further in the context of the Russian analysis. Overall, there is an obvious correlation between 20 vs Zf, with the diagonal running through the two groups. However, there is a great deal of spread from the correlation line within each group.

Breakup altitudes. The quantitative analysis of the EN-PN data is discussed in conjunction with the Russian analysis program, but some qualitative observations are in order. The 10 m daylight Earth-grazing object of 10 August is the largest object in Table I, but at magnitude -19, it is not the brightest by two magnitudes. That is because it was slower than other objects and remained at a higher altitude, where radiation is reduced in proportion to density. EN041274 (the Sumava event) was two magnitudes brighter because it was almost twice as fast, even though it was a type IIIB and broke up 15 km higher. PN40503 was one magnitude brighter

Zf- -H In (S/CVo2). (8)

5

- .

than the Earth-grazing object because it was faster and penetrated twice as deep. Conversely, EN170171 (Wurzburg) was two magnitudes dimmer ~ than the Earth-grazing object because at 2.2 m it had only 5% as much area. PN39470 was also about 2 m, but was 50% faster. EN070591 (Benesov) struck vertically to a lower altitude. PN39121 was much dimmer because it was a IIIB that broke up at a very high altitude.

magnitude -19 because it was stony and massive. EN010677 (Freising) was two magnitudes dimmer because it was carbonaceous and lighter. EN100469 (Otterskir) was another two magnitudes dimmer because it entered 40% slower. The carbonaceous PN39434 and MOFW925 (G. Prairie) had comparable sizes, masses, and magnitudes. The 15 km lower altitude of PN39434 compensated for the two-fold higher velocity of G. Prairie. The soft cometary PN40401 and EN140863 had comparable sizes, masses, and magnitudes; the latter was distinguished by a greater angle of incidence. The trends in magnitude correlate reasonably well with measured velocities, altitudes, and trajectories and inferred size and strength parameters. However, each of these parameters is subject to dispute. The sizes are uncertain to factors of two; the masses to about an order of magnitude, and the strengths to a like amount. While the European and Prairie Network observations provide a useful data base for large objects, these uncertainties can only be addressed when the sensitivity of predictions to uncertainties in these parameters is assessed, as is done in discussion of the Russian analysis program.

ablation, and radiation of meteor impacts at the level needed to infer meteor parameters from observations. The previous note compared that model to other U.S. models with different but related features and objectives. The most closely related model solves the same equations numerically, uses a consistent treatment of fragmentation, and produces similar results for common parameters, although it does not treat radiation in detail and assumes constant heat transfer coefficients. A second model differs primarily in its estimate of fragment spread velocities and variable heat transfer coefficients. Both are primarily concerned with interpreting

The remaining objects only span the interval from 1.8 to 1 m. EN070459 (Pribram) had

Summary and conclusions. Earlier notes derived a model for the hydrodynamics,

the large Tunguska impactor, so they simplify many features and concentrate on varying parameters such as heat of vaporization and material strength to try to match observed surface damage patterns. The previous note discussed models with zero-dimensional hydrodynamics, simplified radiation transport, constant heat transport, parametric fracture and spread, and limited recourse to the limited data on objects tens to thousands of meters across.

analyses have used simplified models to try to achieve agreement with the very limited data on large objects while the European efforts have used more complex models that use observational data more directly to achieve agreement for a larger body of data on smaller objects. While the

This note extends that discussion to European data and models of meteor impact. U.S.

6

former is at a level of simplification better suited to the model for data inversion needed for this effort, the latter represent a more thorough test of model ~ sensitivities and uncertainties for objects in the size range of interest. Thus, it is also a useful test of the inversion model. Fortunately, as shown below, in cases of mutual applicability, the U.S. and European models appear to give broadly consistent results. This note puts the inversion model into a form appropriate for interpretation of the European and Prairie Network observations of 14 objects larger than 1 m in diameter. Over the last 35 years, the average rate of impact of such objects is about 0.4/yr, smaller for larger objects. The EN and PN observations use largely consistent classifications of these objects into stony, carbonaceous, and cometary materials. EN and PN observations indicate that most objects larger than 1 m are expected to be stony or cometary.

When these objects are displayed according to their observed altitude of maximum brightness, which is related to fragmentation altitude, and inferred strength in bars, they fall into three groups: one stony object, the carbonaceous objects, and the soft cometary objects. Ideally, they should fall along the theoretical line of pressure versus altitude, but in practice they exhibit several order of magnitude scatter due to variations in material strength. Correcting pressure for entry velocity provides a more precise measure of deviations from theory. Observed versus predicted fragmentation altitude indicates two groups of objects. Type I and II objects are predicted to fragment at 35-45 km and are observed to fragment at 20-60 km. IIIB objects are predicted to fragment at 70-80 km and are observed to fragment at 60-90 km due to flaws in the larger objects, which reduce their effective strength. There is a strong correlation, but there is a great deal of spread within each group.

The quantitative analysis of the EN-PN data is better discussed in conjunction with the Russian analysis program, but some qualitative observations are useful. The change in magnitude of the 14 large objects with size and speed can be explained readily by the changes in their size, speed, trajectory, and type. Trends in magnitude correlate reasonably well with measured velocities, altitudes, and trajectories and infemd size and strength parameters. However, each parameter is subject to dispute. The sizes are uncertain to factors of two; the masses to about an order of magnitude, and the strengths to a like amount. Thus, while the European and Prairie Network observations provide a useful data base for large objects, these uncertainties can only be addressed when the sensitivity of predictions to uncertainties in these parameters is assessed.

7

References

1. G. Canavan, "Radiation from Hard Objects," Los Alamos National Laboratory report LA-UR- 97-662, February 1997."

2. G. Canavan, "Deceleration and Radiation of Strong, Hard, Asteroids During Atmospheric Impact," Los Alamos National Laboratory report LA-UR-95-1477,l Feb 1997.

3. G. Canavan, "Meteor Signature Interpretation," Los Alamos National Laboratory report LA- UR-97-128, January 1997.

4. G. Canavan, "Fragmentation and Ablation During Entry,'' Los Alamos National Laboratory report LA-UR-97-, May 1997.

5. G. Canavan, "Comparison of Models of Meteor Impact," Los Alamos National Laboratory report LA-UR-97-, May 1997."

6. J. Hills and M. Goda, "The Fragmentation of Small Asteroids in the Atmosphere," Astronomical Journal, Vol. 105, No. 3, March 1933, pp. 11 14-1144.

7. C. Chyba, P. Thomas, and K. Zahnle, "The 1908 Tunguska explosion: atmospheric disruption of a stony asteroid," Nature, Vol. 361,7 January 1993, pp. 40-44.

8. G. Canavan, "Fragmentation of Weak Non-Ablating Objects During Entry,'' Los Alamos National Laboratory report LA-UR-97-, April 1997.

9. C. Chyba, P. Thomas, and K. Zahnle, "The 1908 Tunguska explosion: atmospheric disruption of a stony asteroid," op. cit.

10. Z. Ceplecha, "Earth's Influx of Different Populations of Sporadic Meteoroids from Photographic and Television Data," Bull. Astron. Inst. Czechosl. 39 (1988) 221-236.

11. Z. Ceplecha, "Impacts of meteoroids larger than 1 m into the Earth's atmosphere," Astron. Astrophys. 286,967-970,1994.

12. Z . Ceplecha, "Earth's Influx of Different Populations of Sporadic Meteoroids from Photographic and Television Data," op. cit.

13. C . Chyba, P. Thomas, and K. Zahnle, "The 1908 Tunguska explosion: atmospheric disruption of a stony asteroid," op. cit.

8

Tablef .Bodies with size over 1 m with data derived from photographic observations of their fireballs. (Data on the "Daylight Earth-Grazing" fireball were derived from infrared tracking).

number Daylight W041274 PN40503 EN170171 PN39470 EN070591 PN39121 name Earth-Grazing Sumava -Wiirzburg BeneSov year date 1972 Aug 10 74Dec4 69Oct9 71Jan17 66Decl l 91May7 65Dec27 time UT 20h29" 17h56.6m 7h16m39' 18h20m lh12m00s 23h03m58a 8h21m055

(Source : Ceplecha Table 2)

size [m] (10) 7.1 2.6 2.2 2.0 1.9 1.9 Mmoz (- 19) -2 1 -20 - 17 -18 - 18.5 -12.1 mm Fgl IO6 5m 17000 1600 8500 7500 lo00 hw,,,,Fml 58 *) 73 23 60 33 26 72 h s Fml 76 99 93 75 79 98 96 hE [km] 1 02 55 21 45 22 16 61 cos ZR 0. 0.460 0.619 0.248 0.559 0.986 0.857 ~m F d s 1 15.08 27.0 21.0 15.7 23.6 21.086 23.8 VE F ~ I 14.21 26. a [AUI 1.661 1.472 1.98

11. 2.02

11. 2.35

9.3 2.1 1

2. 13.6 2.428 2.1 1

e 0.390 0.363 0.76 0.64 0.62 0.71 0.619 0.71 Q [Avl 1.0127 0.9369 0.471 ' 0.722 0.90 1 0.6 13 0.9246 0.597 Q [Avl 2.3 1 2.006 3.5 3.3 3.8 3.7 3.93 3.6 w [degl 355.6 315.76 282.0 73 321.0 265.1 218.65 86.5

WegJ 317.956 317.949 251.78 16 116.75 259 46.314 95 i [degl 15.22 6.93 2.3 12.6 2.71 0.8 23.70 5.4 tY Pe (11) IIIB 11 IIIB I1 n IIIB note infrared 4 GS 2 GS

number EN07C459 EN 010677 EN100469 PN39434 MORP 925 PN40401 EN140863 name Phiram Freising Otterskir. G. Prairie year date 59 Apr 7 77 June 1 69 Apr 10 66 Nov 5 84 Feb 23 69 June 29 63 Aug 14 time UT 19h30m215 21h46m 21h44Sm 10h41ml18 2h16m 9h25m135 21h48m345 size [m] 1.8 1.4 1.3 1.3 1.2 1.1 1 .o Mmaz -19.2 - 17 -15.4 - 14.2 - 14.9 -14.0 -15

h.h.irnarFm1 46 41 45 32 48 92 82 hB Fm1 9s 78 84 77 91 104 103 h~ Fml 13 27 24 27 (30) 72 69 cos ZR 0.678 0.328 0.569 0.512 x 0.1 0.500 0.948 v, [kds] 20.886 27.0 16.1 14.6 26.5 19.2 24.6 V E [kml (3) 5.2 G 3 . = 7 (4) 17.5 = 24 a [AUl 2.401 1.3 2.32 0.71 1.93 3.80 3.01 e 0.671 0.75 0.60 0.43 0.73 0.76 0.67 Q [Avl 0.7894 0.455 0.934 0.405 0.5 15 0.907 1.006 Q [Avl 4.0 1 3.2 3.7 1.02 3.35 6.7 5.0 " tdegl 241.75 286 35.1 196.7 97.6 221.3 190.31 0 [degl 17.110 71.01 200.561 42 153.18 97 141.192 i Wegl 10.48 1.5 6.77 3.4 8.4 13.1 35.1 type I 11 I1 11 I1 IIIB IIIB note meteorites PS

tracking 28-5 k m m 67-22bJmm

mco Fg] 11000 2600 2500 2100 2000 **) 200 140

recovered at perigee photometric mass from initial size before entering the atmosphere maximum absolute (100 km distance) magnitude initial mass transformed to the same corrected scale as used in papers of Ceplecha (1988. 1992) height at the maximum brightness height at the beginning and at the end of the luminous trajectory, respectively cosine of the zenith distance of the radiant initial velocity before entering the atmosphere

Idt = 7.07 x 10' given by Halliday (1985)

velocity at the end of the luminous trajectory orbital elements (1950.0) GS ... grating spectra available PS ... prism spectrum available

00 L- a US A n a . #... n A a. A A 9 n e ut3 U L U Y U-l u v UZ U C U C

t I I I I I I I I I I I

P r

100'0

10'0

1'0

1

0 1

00 1

000 1

8

u3 0

0 0

Ln

I 0

0

i .

4

aJ a, h 5 u-l 0

?

# >

0

0 v,

0 d

0 cu

0 Y